PD-1/PD-L1抑制剂治疗晚期胃癌或胃食管结合部肿瘤的疗效和安全性Meta分析
A Meta-Analysis of Efficacy and Safety of PD-1/PD-L1 Inhibitors in the Treatment of Advanced Gastric Cancer or Gastroesophageal Junction Cancer

目的：系统地评价PD-1/PD-L1抑制剂对于晚期阶段胃癌或者胃食管结合部肿瘤(Gastric Cancer or Gastroesophageal Junction Cancer, GC/GEJC)的应用效果及安全性，以期为临床上该疾病的治疗提供决策依据。方法：通过检索中英文的各大数据库文献，如万方、CNKI、维普、Pub Med、Clinical Trials、EMBASE以及Cochrane Library，检索近十余年公开途径发表的PD-1/PD-L1抑制剂用在GC/GEJC治疗中的相关性文献，依照入选标准选择符合的随机对照试验(Randomized Controlled Trial, RCT)，进行数据提取环节并评价纳入研究的偏倚风险后，以软件Review Manager 5.3对结局指标予以分析。结果：纳入的8篇文献合计6345例的患者，其中的PD-1/PD-L1抑制剂治疗(试验组)有3257例的病例，常规治疗(对照组)有3088例的病例。全部是PD-1/PD-L1抑制剂单药或联合化疗与常规化疗方案或安慰剂的总生存期(Overall Survival, OS)对比。相较于对照组，试验组的治疗方案可延长晚期GC/GEJC患者OS (HR = 0.79, 95% CI: 0.74~0.84, P < 0.0001)。亚组分析提示，PD-1/PD-L1抑制剂单药对比化疗或安慰剂、PD-1/PD-L1抑制剂联合化疗对比化疗均能够改善晚期GC/GEJC患者的OS；PD-1/PD-L1抑制剂的给药方案不能改善患者的PFS (HR = 0.94, 95% CI: 0.74~1.19, P = 0.59)。亚组分析结果提示，PD-1/PD-L1抑制剂单药与化疗组相比并不能改善患者PFS (HR = 1.28, 95% CI: 1.05~1.56, P = 0.01)；PD-1/PD-L1抑制剂单药对比安慰剂、PD-1/PD-L1抑制剂联合化疗对比化疗组能提高患者PFS (HR = 0.73, 95% CI: 0.67~0.80, P < 0.00001)；综合分析结果提示PD-1/PD-L1抑制剂不能改善患者客观缓解率(RR = 1.0, 95% CI: 0.78~1.29, P = 0.97)。亚组分析结果显示，PD-1/PD-L1抑制剂联合化疗与常规化疗相比、PD-1/PD-L1抑制剂单药与安慰剂相比均能改善患者的客观缓解率。但PD-1/PD-L1抑制剂单药与常规化疗相比并无优势结果提示试验组引发的治疗相关性的不良事件偏少，安全性更高(RR = 0.85, 95% CI: 0.74~0.99, P = 0.03)。亚组分析结果提示PD-1/PD-L1抑制剂单药与化疗相比，治疗相关不良事件发生率低(RR = 0.69, 95% CI: 0.60~0.81, P < 0.00001)；PD-1/PD-L1抑制剂单药与安慰剂相比、PD-1/PD-L1抑制剂联合化疗与化疗相比较，安全性并没有提高；8篇入选文献以OS、PFS指标测值作为结局目标，对其实施偏倚性的分析，漏斗图显示图形在两侧基本对称，无明显发表偏倚。结论：PD-1/PD-L1抑制剂联合化疗与常规化疗相比可以改善晚期GC/GEJC患者的OS、PFS及客观缓解率，但并没有提高安全性。PD-1/PD-L1抑制剂单药对比常规化疗，能改善晚期GC/GEJC患者OS，且有更高的安全性，但不能提高患者的PFS。
Objective: Systematically evaluate the efficacy and safety of PD-1/PD-L1 inhibitors in the treatment of advanced gastric cancer or gastroesophageal junction cancer (GC/GEJC), to provide a decision-making basis for the clinical treatment of this disease. Methods: By searching major databases in both Chinese and English, such as Wanfang, CNKI, VIP, Pub Med, Clinical Trials, EMBASE, and Cochrane Library, we searched for relevant literature on the use of PD-1/PD-L1 inhibitors in GC/GEJC treatment published through public channels in the past decade. According to the inclusion criteria, we selected randomized controlled trial (RCT) that met the criteria, conducted data extraction, and evaluated the risk of bias included in the study. After that, we analyzed the outcome indicators using the software RevMan 5.3. Results: A total of 6345 patients were included in the eight studies, of which 3257 cases were treated with PD-1/PD-L1 inhibitors (experimental group) and 3088 cases were treated with conventional treatment (control group). Overall survival (OS) was compared between PD-1/PD-L1 inhibitor monotherapy or combination